Delft University of Technology

The primary goal of pattern recognition is supervised or unsupervised classification. Among the various frameworks in which pattern recognition has been traditionally formulated, the statistical approach has been most intensively studied and used in practice. More recently, neural network techniques and methods imported from statistical learning theory have been receiving increasing attention. The design of a recognition system requires careful attention to the following issues: definition of pattern classes, sensing environment, pattern representation, feature extraction and selection, cluster analysis, classifier design and learning, selection of training and test samples, and performance evaluation. In spite of almost 50 years of research and development in this field, the general problem of recognizing complex patterns with arbitrary orientation, location, and scale remains unsolved. New and emerging applications, such as data mining, web searching, retrieval of multimedia data, face recognition, and cursive handwriting recognition, require robust and efficient pattern recognition techniques. The objective of this review paper is to summarize and compare some of the well-known methods used in various stages of a pattern recognition system and identify research topics and applications which are at the forefront of this exciting and challenging field.

We develop a common theoretical framework for combining classifiers which use distinct pattern representations and show that many existing schemes can be considered as special cases of compound classification where all the pattern representations are used jointly to make a decision. An experimental comparison of various classifier combination schemes demonstrates that the combination rule developed under the most restrictive assumptions-the sum rule-outperforms other classifier combinations schemes. A sensitivity analysis of the various schemes to estimation errors is carried out to show that this finding can be justified theoretically.

Designers have been moving increasingly closer to the future users of what they design and the next new thing in the changing landscape of design research has become co-designing with your users. But co-designing is actually not new at all, having taken distinctly different paths in the US and in Europe. The evolution in design research from a user-centred approach to co-designing is changing the roles of the designer, the researcher and the person formerly known as the ‘user’. The implications of this shift for the education of designers and researchers are enormous. The evolution in design research from a user-centred approach to co-designing is changing the landscape of design practice as well, creating new domains of collective creativity. It is hoped that this evolution will support a transformation toward more sustainable ways of living in the future.

Majorana fermions are particles identical to their own antiparticles. They have been theoretically predicted to exist in topological superconductors. Here, we report electrical measurements on indium antimonide nanowires contacted with one normal (gold) and one superconducting (niobium titanium nitride) electrode. Gate voltages vary electron density and define a tunnel barrier between normal and superconducting contacts. In the presence of magnetic fields on the order of 100 millitesla, we observe bound, midgap states at zero bias voltage. These bound states remain fixed to zero bias, even when magnetic fields and gate voltages are changed over considerable ranges. Our observations support the hypothesis of Majorana fermions in nanowires coupled to superconductors.

As far back as the industrial revolution, significant development in technical innovation has succeeded in transforming numerous manual tasks and processes that had been in existence for decades where humans had reached the limits of physical capacity. Artificial Intelligence (AI) offers this same transformative potential for the augmentation and potential replacement of human tasks and activities within a wide range of industrial, intellectual and social applications. The pace of change for this new AI technological age is staggering, with new breakthroughs in algorithmic machine learning and autonomous decision-making, engendering new opportunities for continued innovation. The impact of AI could be significant, with industries ranging from: finance, healthcare, manufacturing, retail, supply chain, logistics and utilities, all potentially disrupted by the onset of AI technologies. The study brings together the collective insight from a number of leading expert contributors to highlight the significant opportunities, realistic assessment of impact, challenges and potential research agenda posed by the rapid emergence of AI within a number of domains: business and management, government, public sector, and science and technology. This research offers significant and timely insight to AI technology and its impact on the future of industry and society in general, whilst recognising the societal and industrial influence on pace and direction of AI development.

We present the final data release of observations of λ21-cm emission from Galactic neutral hydrogen over the entire sky, merging the Leiden/Dwingeloo Survey (LDS: Hartmann & Burton 1997, Atlas of Galactic Neutral Hydrogen) of the sky north of δ = -30° with the Instituto Argentino de Radioastronomía Survey (IAR: Arnal et al. 2000, A&AS, 142, 35; and Bajaja et al. 2005, A&A, 440, 767) of the sky south of δ = -25°. The angular resolution of the combined material is HPBW ~ . The LSR velocity coverage spans the interval -450 km s-1 to +400 km s-1, at a resolution of 1.3 km s-1. The data were corrected for stray radiation at the Institute for Radioastronomy of the University of Bonn, refining the original correction applied to the LDS. The rms brightness-temperature noise of the merged database is K. Residual errors in the profile wings due to defects in the correction for stray radiation are for most of the data below a level of mK. It would be necessary to construct a telescope with a main beam efficiency of % to achieve the same accuracy. The merged and refined material entering the LAB Survey of Galactic H i is intended to be a general resource useful to a wide range of studies of the physical and structural characteristices of the Galactic interstellar environment. The LAB Survey is the most sensitive Milky Way H i survey to date, with the most extensive coverage both spatially and kinematically.

A new energy paradigm, consisting of greater reliance on renewable energy sources and increased concern for energy efficiency in the total energy lifecycle, has accelerated research into energy-related technologies. Due to their ubiquity, magnetic materials play an important role in improving the efficiency and performance of devices in electric power generation, conditioning, conversion, transportation, and other energy-use sectors of the economy. This review focuses on the state-of-the-art hard and soft magnets and magnetocaloric materials, with an emphasis on their optimization for energy applications. Specifically, the impact of hard magnets on electric motor and transportation technologies, of soft magnetic materials on electricity generation and conversion technologies, and of magnetocaloric materials for refrigeration technologies, are discussed. The synthesis, characterization, and property evaluation of the materials, with an emphasis on structure-property relationships, are discussed in the context of their respective markets, as well as their potential impact on energy efficiency. Finally, considering future bottlenecks in raw materials, options for the recycling of rare-earth intermetallics for hard magnets will be discussed.

Complex networks arise in a wide range of biological and sociotechnical systems. Epidemic spreading is central to our understanding of dynamical processes in complex networks, and is of interest to physicists, mathematicians, epidemiologists, and computer and social scientists. This review presents the main results and paradigmatic models in infectious disease modeling and generalized social contagion processes.

The transition within business from a linear to a circular economy brings with it a range of practical challenges for companies. The following question is addressed: What are the product design and business model strategies for companies that want to move to a circular economy model? This paper develops a framework of strategies to guide designers and business strategists in the move from a linear to a circular economy. Building on Stahel, the terminology of slowing, closing, and narrowing resource loops is introduced. A list of product design strategies, business model strategies, and examples for key decision-makers in businesses is introduced, to facilitate the move to a circular economy. This framework also opens up a future research agenda for the circular economy.

Transformative artificially intelligent tools, such as ChatGPT, designed to generate sophisticated text indistinguishable from that produced by a human, are applicable across a wide range of contexts. The technology presents opportunities as well as, often ethical and legal, challenges, and has the potential for both positive and negative impacts for organisations, society, and individuals. Offering multi-disciplinary insight into some of these, this article brings together 43 contributions from experts in fields such as computer science, marketing, information systems, education, policy, hospitality and tourism, management, publishing, and nursing. The contributors acknowledge ChatGPT’s capabilities to enhance productivity and suggest that it is likely to offer significant gains in the banking, hospitality and tourism, and information technology industries, and enhance business activities, such as management and marketing. Nevertheless, they also consider its limitations, disruptions to practices, threats to privacy and security, and consequences of biases, misuse, and misinformation. However, opinion is split on whether ChatGPT’s use should be restricted or legislated. Drawing on these contributions, the article identifies questions requiring further research across three thematic areas: knowledge, transparency, and ethics; digital transformation of organisations and societies; and teaching, learning, and scholarly research. The avenues for further research include: identifying skills, resources, and capabilities needed to handle generative AI; examining biases of generative AI attributable to training datasets and processes; exploring business and societal contexts best suited for generative AI implementation; determining optimal combinations of human and generative AI for various tasks; identifying ways to assess accuracy of text produced by generative AI; and uncovering the ethical and legal issues in using generative AI across different contexts.

Abstract Classical least squares regression consists of minimizing the sum of the squared residuals. Many authors have produced more robust versions of this estimator by replacing the square by something else, such as the absolute value. In this article a different approach is introduced in which the sum is replaced by the median of the squared residuals. The resulting estimator can resist the effect of nearly 50% of contamination in the data. In the special case of simple regression, it corresponds to finding the narrowest strip covering half of the observations. Generalizations are possible to multivariate location, orthogonal regression, and hypothesis testing in linear models.

This article reports the findings of AI4People, an Atomium-EISMD initiative designed to lay the foundations for a "Good AI Society". We introduce the core opportunities and risks of AI for society; present a synthesis of five ethical principles that should undergird its development and adoption; and offer 20 concrete recommendations-to assess, to develop, to incentivise, and to support good AI-which in some cases may be undertaken directly by national or supranational policy makers, while in others may be led by other stakeholders. If adopted, these recommendations would serve as a firm foundation for the establishment of a Good AI Society.

Ballistic point contacts, defined in the two-dimensional electron gas of a GaAs-AlGaAs heterostructure, have been studied in zero magnetic field. The conductance changes in quantized steps of ${e}^{2}$/\ensuremath{\pi}\ensuremath{\Elzxh} when the width, controlled by a gate on top of heterojunction, is varied. Up to sixteen steps are observed when the point contact is widened from 0 to 360 nm. An explanation is proposed, which assumes quantized transverse momentum in the point-contact region.

Here we provide global estimates of the seasonal flux of sediment, on a river-by-river basis, under modern and prehuman conditions. Humans have simultaneously increased the sediment transport by global rivers through soil erosion (by 2.3 +/- 0.6 billion metric tons per year), yet reduced the flux of sediment reaching the world's coasts (by 1.4 +/- 0.3 billion metric tons per year) because of retention within reservoirs. Over 100 billion metric tons of sediment and 1 to 3 billion metric tons of carbon are now sequestered in reservoirs constructed largely within the past 50 years. African and Asian rivers carry a greatly reduced sediment load; Indonesian rivers deliver much more sediment to coastal areas.

In this tutorial review, an overview of the why, what and how of enzyme immobilisation for use in biocatalysis is presented. The importance of biocatalysis in the context of green and sustainable chemicals manufacture is discussed and the necessity for immobilisation of enzymes as a key enabling technology for practical and commercial viability is emphasised. The underlying reasons for immobilisation are the need to improve the stability and recyclability of the biocatalyst compared to the free enzyme. The lower risk of product contamination with enzyme residues and low or no allergenicity are further advantages of immobilised enzymes. Methods for immobilisation are divided into three categories: adsorption on a carrier (support), encapsulation in a carrier, and cross-linking (carrier-free). General considerations regarding immobilisation, regardless of the method used, are immobilisation yield, immobilisation efficiency, activity recovery, enzyme loading (wt% in the biocatalyst) and the physical properties, e.g. particle size and density, hydrophobicity and mechanical robustness of the immobilisate, i.e. the immobilised enzyme as a whole (enzyme + support). The choice of immobilisate is also strongly dependent on the reactor configuration used, e.g. stirred tank, fixed bed, fluidised bed, and the mode of downstream processing. Emphasis is placed on relatively recent developments, such as the use of novel supports such as mesoporous silicas, hydrogels, and smart polymers, and cross-linked enzyme aggregates (CLEAs).

In this paper we describe T-MAC, a contention-based Medium Access Control protocol for wireless sensor networks. Applications for these networks have some characteristics (low message rate, insensitivity to latency) that can be exploited to reduce energy consumption by introducing an activesleep duty cycle. To handle load variations in time and location T-MAC introduces an adaptive duty cycle in a novel way: by dynamically ending the active part of it. This reduces the amount of energy wasted on idle listening, in which nodes wait for potentially incoming messages, while still maintaining a reasonable throughput.We discuss the design of T-MAC, and provide a head-to-head comparison with classic CSMA (no duty cycle) and S-MAC (fixed duty cycle) through extensive simulations. Under homogeneous load, T-MAC and S-MAC achieve similar reductions in energy consumption (up to 98%) compared to CSMA. In a sample scenario with variable load, however, T-MAC outperforms S-MAC by a factor of 5. Preliminary energy-consumption measurements provide insight into the internal workings of the T-MAC protocol.

The canonical example of a quantum-mechanical two-level system is spin. The simplest picture of spin is a magnetic moment pointing up or down. The full quantum properties of spin become apparent in phenomena such as superpositions of spin states, entanglement among spins, and quantum measurements. Many of these phenomena have been observed in experiments performed on ensembles of particles with spin. Only in recent years have systems been realized in which individual electrons can be trapped and their quantum properties can be studied, thus avoiding unnecessary ensemble averaging. This review describes experiments performed with quantum dots, which are nanometer-scale boxes defined in a semiconductor host material. Quantum dots can hold a precise but tunable number of electron spins starting with 0, 1, 2, etc. Electrical contacts can be made for charge transport measurements and electrostatic gates can be used for controlling the dot potential. This system provides virtually full control over individual electrons. This new, enabling technology is stimulating research on individual spins. This review describes the physics of spins in quantum dots containing one or two electrons, from an experimentalist's viewpoint. Various methods for extracting spin properties from experiment are presented, restricted exclusively to electrical measurements. Furthermore, experimental techniques are discussed that allow for (1) the rotation of an electron spin into a superposition of up and down, (2) the measurement of the quantum state of an individual spin, and (3) the control of the interaction between two neighboring spins by the Heisenberg exchange interaction. Finally, the physics of the relevant relaxation and dephasing mechanisms is reviewed and experimental results are compared with theories for spin-orbit and hyperfine interactions. All these subjects are directly relevant for the fields of quantum information processing and spintronics with single spins (i.e., single spintronics).

This paper presents an overview of integrated optics routing and coupling devices based on multimode interference. The underlying self-imaging principle in multimode waveguides is described using a guided mode propagation analysis. Special issues concerning the design and operation of multimode interference devices are discussed, followed by a survey of reported applications. It is shown that multimode interference couplers offer superior performance, excellent tolerance to polarization and wavelength variations, and relaxed fabrication requirements when compared to alternatives such as directional couplers, adiabatic X- or Y-junctions, and diffractive star couplers.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

The chemical industry is under considerable pressure to replace many of the volatile organic compounds (VOCs) that are currently used as solvents in organic synthesis. The toxic and/or hazardous properties of many solvents, notably chlorinated hydrocarbons, combined with serious environmental issues, such as atmospheric emissions and contamination of aqueous effluents is making their use prohibitive. This is an important driving force in the quest for novel reaction media. Curzons and coworkers, for example, recently noted that rigorous management of solvent use is likely to result in the greatest improvement towards greener processes for the manufacture of pharmaceutical intermediates. The current emphasis on novel reaction media is also motivated by the need for efficient methods for recycling homogeneous catalysts. The key to waste minimisation in chemicals manufacture is the widespread substitution of classical 'stoichiometric' syntheses by atom efficient, catalytic alternatives. In the context of homogeneous catalysis, efficient recycling of the catalyst is a conditio sine qua non for economically and environmentally attractive processes. Motivated by one or both of the above issues much attention has been devoted to homogeneous catalysis in aqueous biphasic and fluorous biphasic systems as well as in supercritical carbon dioxide. Similarly, the use of ionic liquids as novel reaction media may offer a convenient solution to both the solvent emission and the catalyst recycling problem.

Millimeter wave (mmWave) cellular systems will enable gigabit-per-second data rates thanks to the large bandwidth available at mmWave frequencies. To realize sufficient link margin, mmWave systems will employ directional beamforming with large antenna arrays at both the transmitter and receiver. Due to the high cost and power consumption of gigasample mixed-signal devices, mmWave precoding will likely be divided among the analog and digital domains. The large number of antennas and the presence of analog beamforming requires the development of mmWave-specific channel estimation and precoding algorithms. This paper develops an adaptive algorithm to estimate the mmWave channel parameters that exploits the poor scattering nature of the channel. To enable the efficient operation of this algorithm, a novel hierarchical multi-resolution codebook is designed to construct training beamforming vectors with different beamwidths. For single-path channels, an upper bound on the estimation error probability using the proposed algorithm is derived, and some insights into the efficient allocation of the training power among the adaptive stages of the algorithm are obtained. The adaptive channel estimation algorithm is then extended to the multi-path case relying on the sparse nature of the channel. Using the estimated channel, this paper proposes a new hybrid analog/digital precoding algorithm that overcomes the hardware constraints on the analog-only beamforming, and approaches the performance of digital solutions. Simulation results show that the proposed low-complexity channel estimation algorithm achieves comparable precoding gains compared to exhaustive channel training algorithms. The results illustrate that the proposed channel estimation and precoding algorithms can approach the coverage probability achieved by perfect channel knowledge even in the presence of interference.